The emergence of variants of HIV PR resistant to small-molecule PR inhibitors is well documented. While strategies involving combinations of inhibitors may eventually block the emergence of resistant variants, we have pursued an alternative approach to protease inhibition that may be less sensitive to point mutations. Molecular modeling is used to design defective HIV PR monomers which exert a dominant-negative (d-n) inhibitory effect, presumably leading to the formaion of catalytically defective protease heterodimers in vivo, and ultimately yielding non-infectious viral particles. A PR variant containing three amino acid changes D25K, G49W and I50W (KWW) was the most effective d-n inhibitor assayed in cultured cells. Gene therapy vectors, in combination with potent vector-producing cell lines, have been successfully used to deliver our KWW PR inhibitor to PBMCs along with other d-n inhibitors of HIV-1 replication (e.g., RevM10). Virus challenge assays of these cells shows that up to a 1.5 log protection is afforded by the expression of the KWW variant (against HIV-1 IIIB). Additional virus challenge assays are in progress. Our goal is to develop the HIV PR d-n inhibitor as a potential gene therapy tool while using HIV PR as a model system for the study of protein-protein interactions and to gain an understanding of dominant-negative interference of protein function. We are also using the yeast two-hybrid system to characterize the strength of interaction between wt and variant HIV PR monomers to better understand the mechanism of dominant-negative HIV-1 PR inhibition and the protein-protein interactions responsible for this inhibition in addition to developing our macromolecular PR inhibitors for potential application in vivo.